Toxins 2015, 7, 3903-3915; doi:10.3390/toxins7103903 OPEN ACCESS
toxins ISSN 2072-6651 www.mdpi.com/journal/toxins Article
Development and Evaluation of Monoclonal Antibodies for Paxilline Chris M. Maragos Bacterial Foodborne Pathogens and Mycology Research Unit, USDA-ARS-NCAUR, 1815 North University Street, Peoria, IL 61604, USA; E-Mail: [email protected]; Tel.: +1-309-681-6266; Fax: +1-309-681-6672 Academic Editor: Laura Anfossi Received: 14 August 2015 / Accepted: 22 September 2015 / Published: 25 September 2015
Abstract: Paxilline (PAX) is a tremorgenic mycotoxin that has been found in perennial ryegrass infected with Acremonium lolii. To facilitate screening for this toxin, four murine monoclonal antibodies (mAbs) were developed. In competitive indirect enzyme-linked immunosorbent assays (CI-ELISAs) the concentrations of PAX required to inhibit signal development by 50% (IC50s) ranged from 1.2 to 2.5 ng/mL. One mAb (2-9) was applied to the detection of PAX in maize silage. The assay was sensitive to the effects of solvents, with 5% acetonitrile or 20% methanol causing a two-fold or greater increase in IC50. For analysis of silage samples, extracts were cleaned up by adsorbing potential matrix interferences onto a solid phase extraction column. The non-retained extract was then diluted with buffer to reduce solvent content prior to assay. Using this method, the limit of detection for PAX in dried silage was 15 µg/kg and the limit of quantification was 90 µg/kg. Recovery from samples spiked over the range of 100 to 1000 µg/kg averaged 106% ± 18%. The assay was applied to 86 maize silage samples, with many having detectable, but none having quantifiable, levels of PAX. The results suggest the CI-ELISA can be applied as a sensitive technique for the screening of PAX in maize silage. Keywords: paxilline; tremorgen; antibody; immunoassay; mycotoxin; silage
1. Introduction The tremorgens are a class of fungal secondary metabolites (mycotoxins), which, like the ergot alkaloids, contain a substituted indole-moiety. There are many examples of tremorgens, including
Toxins 2015, 7
3904
the penitrems (tremortins), janthitrems, lolitrems, aflatrem, paxilline, paspalines, paspalitrems, fumitremorgens, tryptoquivalines, verruculogen, and others [1]. They are produced by certain species of Aspergillus, Penicillium, Claviceps, and Acremonium [2–5]. Paxilline (PAX, Figure 1) was first isolated as a metabolite of Penicillium paxilli (P. paxilli) that caused sustained tremors and hypersensitivity to sound when given orally to one-day old cockerels and intra-peritoneally to mice [6]. PAX is a potent inhibitor of high conductance calcium-activated potassium channels [7,8], and has profound effects on the electromyographic activity of smooth muscle of the reticulorumen of sheep [9]. Lolitrem B, a related tremorgen, also inhibits such channels and this effect is believed to be the mechanism of action for the disease “ryegrass staggers” [10]. Paspaline, paspalicine, and paspalinine are structurally similar to PAX [11]. The group is variously known as the paspaline group [2], paspalitrem group [3], or the paspalanes [1]. Members of the group have been associated with Dallisgrass poisoning (aka “paspalum staggers”) in grazing cattle [11]. In addition to neurotoxicity, PAX has been shown to be genotoxic and to cause DNA damage in human lymphocytes [12,13].
Figure 1. Structures of paxilline (MW 435.56) and paxitriol (MW 437.57). PAX was isolated from P. paxilli obtained from insect-damaged pecans [6] and has been found in moldy tomatoes [14]. However, it is the occurrence of lolitrem B and PAX in perennial ryegrass that has been associated with disease, in particular “ryegrass staggers” in livestock [15]. PAX was found at significant levels (up to 14 mg/kg) in samples of ryegrass [16]. It is produced through the intermediate 3-geranylgeranylindole [17], and the biosynthesis has been recently summarized [18]. The chemistry of many of the tremorgens as well as their structural and biosynthetic relationships were well summarized by Cole [3] and Sings and Singh [1]. Perhaps because of the large number of possible tremorgens and the lack of widely available analytical standards for most of them, the development of analytical methods for this group of mycotoxins has not progressed to the same extent as for other common mycotoxins. Most of the early methods for detection of the paspalitrem-type mycotoxins were based upon liquid/liquid partitioning followed by thin layer chromatography (TLC), as summarized by Selala et al. [19]. PAX absorbs in the ultraviolet (UV) region and in methanol (MeOH) demonstrates absorption bands at 230 nm (ε 41,500) and 281 nm (ε 8000) [6]. For this reason, liquid chromatographic (LC) methods have incorporated UV or diode array detectors [14,19,20]. Upon exposure to UV light, PAX yields uncharacterized fluorescent products, with excitation maximum at 360 nm and emission maximum at 462 nm, which suggests LC with fluorescence detection is also possible [21]. Liquid chromatography with mass spectrometric detection (LC-MS) has been used to detect PAX in perennial ryegrass [22]. More recently, LC combined with high resolution MS has been applied to determine the idole-diterpenoid profiles of
Toxins 2015, 7
3905
certain Claviceps species [23]. A screening assay for 186 fungal and bacterial metabolites in indoor matrices using LC with electrospray tandem ionization mass spectrometry (LC-MS/MS) also included PAX [24]. Antibodies for PAX were developed by AgResearch in New Zealand in the 1990s [16]. The antibodies were applied in enzyme-linked immunosorbent assays (ELISAs), and in combinations of TLC and LC with immunochemical detection [16,25,26]. Those appear to be the only published reports of such assays, although recently a commercial biosensor array has incorporated PAX, with a limit of detection (LOD) of 50 µg/kg [27]. Unfortunately, additional specifics of that assay have not been published. The objectives of our research were to develop antibodies and immunoassays for PAX and apply them towards a small-scale survey of PAX in maize silages. 2. Results and Discussion 2.1. Production of mAbs to PAX Ten mice were immunized with a conjugate of paxitriol-hemiglutarate and ovalbumin (RPAX-OVA). Sera were evaluated with a competitive indirect ELISA (CI-ELISA). In this format an immobilized paxilline-bovine serum albumin (PAX-BSA) conjugate competed with free PAX for PAX antibodies. Two of the immunized mice were selected for splenocyte fusions and a total of 15 PAX-responsive cultures were obtained. From these, four antibody-producing monoclonal cell lines were isolated. These were designated mAb 1-4 (isotype IgG1κ), 2-2 (IgG1λ), 2-8 (IgG1λ), and 2-9 (IgG2ακ). Responses of these mAbs in competitive indirect enzyme-linked immunosorbent assays (CI-ELISAs) are depicted in Figure 2.
Figure 2. Response of four mAbs in CI-ELISA of PAX. Data shown are the averages of six plates ±1 standard deviation (SD). Calibration curves of PAX in phosphate buffered saline (PBS) were used to determine the concentrations needed to inhibit color development by 20% (IC20), 50% (IC50) and 80% (IC80) (Table 1). The response curves of mAbs 2-2 and 2-9 were essentially superimposable. Although the response curves were
Toxins 2015, 7
3906
similar, the antibodies were distinctly different, as they had different isotypes (IgG1λ and IgG2ακ). While not as sensitive to PAX, the shapes of the curves from mAbs 1-4 and 2-8 had steeper slopes. For quantitative assays, this resulted in a lower dynamic range for the assays with these two antibodies. A widely used measure of dynamic range for competitive immunoassays is the range of concentrations between the IC20 (minimum) and IC80 (maximum). Table 1. Response parameters for four PAX mAbs in CI-ELISA (data from Figure 2). Antibody 1-4 2-2 2-8 2-9
IC20 (ng/mL) 0.6 0.4 0.9 0.4
IC50 (ng/mL) 1.5 1.2 2.5 1.2
IC80 (ng/mL) 2.7 3.6 5.2 3.6
Dynamic Range (ng/mL) 0.6–2.7 0.4–3.6 0.9–5.2 0.4–3.6
Based upon the parameters in Table 1, mAb 2-9 was chosen as the antibody to use for further ELISA development. However, it should be noted that the attributes of mAb 1-4, with a similar IC50 but much steeper dose-response curve (and lower IC80), might make this a better choice of an antibody for a qualitative immunoassay format, such as for a lateral flow device. The sensitivity of these CI-ELISAs compares well to the previous literature. Garthwaite et al. [16] immunized mice with a PAX-O-(carboxymethyl)oxime-keyhole limpet hemocyanin conjugate and developed a CI-ELISA using immobilized PAX-ovalbumin. Several mAbs were isolated, and two were described in the publication. Although the IC50s were not provided, the dynamic ranges were reported. These were 2 to 1000 ng/mL (for mAb M-03/01) and 5 to 500 ng/mL (for mAb M-03/02) [16]. The dynamic ranges of the four mAbs reported herein were not as broad, however the assays were more sensitive (Table 1). 2.2. Effects of Solvents on mAb 2-9 ELISA PAX is typically extracted using halogenated solvents or aqueous mixtures of organic solvents. Because of the toxicity of chlorinated solvents, the use of alternative organic solvents is preferable. Accordingly, the effects of acetonitrile (ACN) and MeOH on the mAb 2-9-based assay were determined. PAX standards were prepared in mixtures of PBS with solvent at different concentrations then assayed by CI-ELISA (Table 2). Both solvents were detrimental to the sensitivity of the assay. However, the antibody appeared to be much more susceptible to the effects of ACN (Figure 3) than to MeOH (Figure 4). One parameter for assessing the impact on the sensitivity is the relative effect of the solvent on the IC50 (e.g., relative response, Table 2). A relative response of 50% indicates that the IC50 has increased by two-fold (sensitivity has diminished two-fold). For ACN this occurred at a concentration below 5%, whereas for MeOH it occurred at a concentration between 10% and 20%. This suggested that MeOH, rather than ACN would be preferred as an extraction solvent. The mAb 2-9 ELISA appears to have tolerance to MeOH that is similar, or slightly less, than that reported previously for mAbs M-03/01 and M-03/02, which tolerated up to 15% [16]. The mAb 2-9 ELISA did not respond to the indole-containing toxins penitrem-A (PEN-A), or cyclopiazonic acid (CPA), or to the ergot alkaloid ergometrine, when tested at concentrations up to 10 µg/mL. The low cross-reactivities (
toxins ISSN 2072-6651 www.mdpi.com/journal/toxins Article
Development and Evaluation of Monoclonal Antibodies for Paxilline Chris M. Maragos Bacterial Foodborne Pathogens and Mycology Research Unit, USDA-ARS-NCAUR, 1815 North University Street, Peoria, IL 61604, USA; E-Mail: [email protected]; Tel.: +1-309-681-6266; Fax: +1-309-681-6672 Academic Editor: Laura Anfossi Received: 14 August 2015 / Accepted: 22 September 2015 / Published: 25 September 2015
Abstract: Paxilline (PAX) is a tremorgenic mycotoxin that has been found in perennial ryegrass infected with Acremonium lolii. To facilitate screening for this toxin, four murine monoclonal antibodies (mAbs) were developed. In competitive indirect enzyme-linked immunosorbent assays (CI-ELISAs) the concentrations of PAX required to inhibit signal development by 50% (IC50s) ranged from 1.2 to 2.5 ng/mL. One mAb (2-9) was applied to the detection of PAX in maize silage. The assay was sensitive to the effects of solvents, with 5% acetonitrile or 20% methanol causing a two-fold or greater increase in IC50. For analysis of silage samples, extracts were cleaned up by adsorbing potential matrix interferences onto a solid phase extraction column. The non-retained extract was then diluted with buffer to reduce solvent content prior to assay. Using this method, the limit of detection for PAX in dried silage was 15 µg/kg and the limit of quantification was 90 µg/kg. Recovery from samples spiked over the range of 100 to 1000 µg/kg averaged 106% ± 18%. The assay was applied to 86 maize silage samples, with many having detectable, but none having quantifiable, levels of PAX. The results suggest the CI-ELISA can be applied as a sensitive technique for the screening of PAX in maize silage. Keywords: paxilline; tremorgen; antibody; immunoassay; mycotoxin; silage
1. Introduction The tremorgens are a class of fungal secondary metabolites (mycotoxins), which, like the ergot alkaloids, contain a substituted indole-moiety. There are many examples of tremorgens, including
Toxins 2015, 7
3904
the penitrems (tremortins), janthitrems, lolitrems, aflatrem, paxilline, paspalines, paspalitrems, fumitremorgens, tryptoquivalines, verruculogen, and others [1]. They are produced by certain species of Aspergillus, Penicillium, Claviceps, and Acremonium [2–5]. Paxilline (PAX, Figure 1) was first isolated as a metabolite of Penicillium paxilli (P. paxilli) that caused sustained tremors and hypersensitivity to sound when given orally to one-day old cockerels and intra-peritoneally to mice [6]. PAX is a potent inhibitor of high conductance calcium-activated potassium channels [7,8], and has profound effects on the electromyographic activity of smooth muscle of the reticulorumen of sheep [9]. Lolitrem B, a related tremorgen, also inhibits such channels and this effect is believed to be the mechanism of action for the disease “ryegrass staggers” [10]. Paspaline, paspalicine, and paspalinine are structurally similar to PAX [11]. The group is variously known as the paspaline group [2], paspalitrem group [3], or the paspalanes [1]. Members of the group have been associated with Dallisgrass poisoning (aka “paspalum staggers”) in grazing cattle [11]. In addition to neurotoxicity, PAX has been shown to be genotoxic and to cause DNA damage in human lymphocytes [12,13].
Figure 1. Structures of paxilline (MW 435.56) and paxitriol (MW 437.57). PAX was isolated from P. paxilli obtained from insect-damaged pecans [6] and has been found in moldy tomatoes [14]. However, it is the occurrence of lolitrem B and PAX in perennial ryegrass that has been associated with disease, in particular “ryegrass staggers” in livestock [15]. PAX was found at significant levels (up to 14 mg/kg) in samples of ryegrass [16]. It is produced through the intermediate 3-geranylgeranylindole [17], and the biosynthesis has been recently summarized [18]. The chemistry of many of the tremorgens as well as their structural and biosynthetic relationships were well summarized by Cole [3] and Sings and Singh [1]. Perhaps because of the large number of possible tremorgens and the lack of widely available analytical standards for most of them, the development of analytical methods for this group of mycotoxins has not progressed to the same extent as for other common mycotoxins. Most of the early methods for detection of the paspalitrem-type mycotoxins were based upon liquid/liquid partitioning followed by thin layer chromatography (TLC), as summarized by Selala et al. [19]. PAX absorbs in the ultraviolet (UV) region and in methanol (MeOH) demonstrates absorption bands at 230 nm (ε 41,500) and 281 nm (ε 8000) [6]. For this reason, liquid chromatographic (LC) methods have incorporated UV or diode array detectors [14,19,20]. Upon exposure to UV light, PAX yields uncharacterized fluorescent products, with excitation maximum at 360 nm and emission maximum at 462 nm, which suggests LC with fluorescence detection is also possible [21]. Liquid chromatography with mass spectrometric detection (LC-MS) has been used to detect PAX in perennial ryegrass [22]. More recently, LC combined with high resolution MS has been applied to determine the idole-diterpenoid profiles of
Toxins 2015, 7
3905
certain Claviceps species [23]. A screening assay for 186 fungal and bacterial metabolites in indoor matrices using LC with electrospray tandem ionization mass spectrometry (LC-MS/MS) also included PAX [24]. Antibodies for PAX were developed by AgResearch in New Zealand in the 1990s [16]. The antibodies were applied in enzyme-linked immunosorbent assays (ELISAs), and in combinations of TLC and LC with immunochemical detection [16,25,26]. Those appear to be the only published reports of such assays, although recently a commercial biosensor array has incorporated PAX, with a limit of detection (LOD) of 50 µg/kg [27]. Unfortunately, additional specifics of that assay have not been published. The objectives of our research were to develop antibodies and immunoassays for PAX and apply them towards a small-scale survey of PAX in maize silages. 2. Results and Discussion 2.1. Production of mAbs to PAX Ten mice were immunized with a conjugate of paxitriol-hemiglutarate and ovalbumin (RPAX-OVA). Sera were evaluated with a competitive indirect ELISA (CI-ELISA). In this format an immobilized paxilline-bovine serum albumin (PAX-BSA) conjugate competed with free PAX for PAX antibodies. Two of the immunized mice were selected for splenocyte fusions and a total of 15 PAX-responsive cultures were obtained. From these, four antibody-producing monoclonal cell lines were isolated. These were designated mAb 1-4 (isotype IgG1κ), 2-2 (IgG1λ), 2-8 (IgG1λ), and 2-9 (IgG2ακ). Responses of these mAbs in competitive indirect enzyme-linked immunosorbent assays (CI-ELISAs) are depicted in Figure 2.
Figure 2. Response of four mAbs in CI-ELISA of PAX. Data shown are the averages of six plates ±1 standard deviation (SD). Calibration curves of PAX in phosphate buffered saline (PBS) were used to determine the concentrations needed to inhibit color development by 20% (IC20), 50% (IC50) and 80% (IC80) (Table 1). The response curves of mAbs 2-2 and 2-9 were essentially superimposable. Although the response curves were
Toxins 2015, 7
3906
similar, the antibodies were distinctly different, as they had different isotypes (IgG1λ and IgG2ακ). While not as sensitive to PAX, the shapes of the curves from mAbs 1-4 and 2-8 had steeper slopes. For quantitative assays, this resulted in a lower dynamic range for the assays with these two antibodies. A widely used measure of dynamic range for competitive immunoassays is the range of concentrations between the IC20 (minimum) and IC80 (maximum). Table 1. Response parameters for four PAX mAbs in CI-ELISA (data from Figure 2). Antibody 1-4 2-2 2-8 2-9
IC20 (ng/mL) 0.6 0.4 0.9 0.4
IC50 (ng/mL) 1.5 1.2 2.5 1.2
IC80 (ng/mL) 2.7 3.6 5.2 3.6
Dynamic Range (ng/mL) 0.6–2.7 0.4–3.6 0.9–5.2 0.4–3.6
Based upon the parameters in Table 1, mAb 2-9 was chosen as the antibody to use for further ELISA development. However, it should be noted that the attributes of mAb 1-4, with a similar IC50 but much steeper dose-response curve (and lower IC80), might make this a better choice of an antibody for a qualitative immunoassay format, such as for a lateral flow device. The sensitivity of these CI-ELISAs compares well to the previous literature. Garthwaite et al. [16] immunized mice with a PAX-O-(carboxymethyl)oxime-keyhole limpet hemocyanin conjugate and developed a CI-ELISA using immobilized PAX-ovalbumin. Several mAbs were isolated, and two were described in the publication. Although the IC50s were not provided, the dynamic ranges were reported. These were 2 to 1000 ng/mL (for mAb M-03/01) and 5 to 500 ng/mL (for mAb M-03/02) [16]. The dynamic ranges of the four mAbs reported herein were not as broad, however the assays were more sensitive (Table 1). 2.2. Effects of Solvents on mAb 2-9 ELISA PAX is typically extracted using halogenated solvents or aqueous mixtures of organic solvents. Because of the toxicity of chlorinated solvents, the use of alternative organic solvents is preferable. Accordingly, the effects of acetonitrile (ACN) and MeOH on the mAb 2-9-based assay were determined. PAX standards were prepared in mixtures of PBS with solvent at different concentrations then assayed by CI-ELISA (Table 2). Both solvents were detrimental to the sensitivity of the assay. However, the antibody appeared to be much more susceptible to the effects of ACN (Figure 3) than to MeOH (Figure 4). One parameter for assessing the impact on the sensitivity is the relative effect of the solvent on the IC50 (e.g., relative response, Table 2). A relative response of 50% indicates that the IC50 has increased by two-fold (sensitivity has diminished two-fold). For ACN this occurred at a concentration below 5%, whereas for MeOH it occurred at a concentration between 10% and 20%. This suggested that MeOH, rather than ACN would be preferred as an extraction solvent. The mAb 2-9 ELISA appears to have tolerance to MeOH that is similar, or slightly less, than that reported previously for mAbs M-03/01 and M-03/02, which tolerated up to 15% [16]. The mAb 2-9 ELISA did not respond to the indole-containing toxins penitrem-A (PEN-A), or cyclopiazonic acid (CPA), or to the ergot alkaloid ergometrine, when tested at concentrations up to 10 µg/mL. The low cross-reactivities (
